SMW4083 ECONOMETRICS INDIVIDUAL ASSIGNMENT DR. WAN NUR RAHINI AZNIE BINTI ZAINUDIN NAME : HUSNA HAYATI BINTI AHMAD MATRIC NUMBER S : 1162369 COURSE : FINANCIAL MATHEMATICSThe casual relationship in this journal is between poverty and household income. The dependent variable is total number of people in poverty in each country across the United States . Firstly, this study uses simple regression model to looks at median household income’s effect on total poverty.
It appeared to be positively correlated. As the total number of people in poverty increases, the median household income of each country represented in U.S. dollars increases , several independent variables are added which are total unemployment, population an d total number of people with less than a high school education in four multiple regression model. total_allage =?0 + ?1med_house + u ?total _allage ?med _house = 0+ ?1 + 0 ?total _allage ?med _house = ?1 In both simple and multiple regression model, it sho ws the positively correlated . As the number of people in poverty level increase, the median household income increase. Adding others variable such as population also shows the positive correlation between the variables. For this journ al, it uses cross -sectional data.
As shown in the stated figure in the journal, both variable (total number of people in poverty and median household income) is studied and measured. It uses county -level data in the United States . Apart from that, this data set is tak en at a specified period which is in 2013 .
In this study, zero conditional mean assumption ( ;#55349;;#56421;|;#55349;;#56418; =0) applies to this econometrics model and is not violated. This proven when the data is transformed in the form of histogram. It shows a slightly leftwa rd skewness. Other factor that I can think of that may influence the ‘effect variable’ is epidemic diseases such as AIDS and malaria . It will affect the poverty leve l in the country. Next, low productivity in agriculture. As there is low productivity, the required number of workers become low .
So that it affected the poverty level.Simple regression model: Total_allage = -6354.315 + 0.387med_house + u ?total _allage ?med _house = 0+ 0.387 + 0 ?total _allage ?med _house = 0.387 In this simple regression model, the sign is positive which means that i t has positive relationship between total number of people in poverty and median household income. The coefficient for ?med _house is 0.
387. It means that as the median household income increases by one dollar, the total number of people in poverty increases by 0.387. Multiple regression model: total_allage = 1735.92 + 0.
105 med_house + 0.0562 population + 0.0101 tot_unemp +0.0053 educ_less_high N= 3100 R-squared= 0.3797 In this multiple regression model, median household income shows positive effect on total number of people in poverty. It shows the positive correlated relationship between all the variables.
R -squared for this model is 0.3797. It means 37.97% of variation in y (total number of people in po verty) is caused by variation in all explanatory variables in the model. Simple regression model: Total_allage =?0 + ?1med_house + u Total_allage = -6354.
315 + 0.387med_house + u (1551.349 ) ( 0.
0327 ) N= 3111 t-statistic= 11.83 R-squared= 0.0431 t-critical= 2.
33 It shows positive correlation between median household income and total number of people in poverty. As median household income increases by one dollar, the total number of people in poverty increases by 0.387. For standard normal distribution test, t-statistic more than critical value thus it statistically significant at 1% of significant level .total_allage = ?0 + ?1med_house + ?2population + u 1763.588 0.107 0.
0563 (1266.65 ) ( 0.0272 ) ( 0.0014 ) t-statistic= 1.39 3.
94 40.92 t-critical= 2.33 2.33 2.33 N= 3110 R-squared= 0.3783 Median household income and population estimate s of each country is statistically significant at 1% of significant level. Thus, it shows the positive correlation.
total_allage = ?0 + ?1med_house + ?2population + ?3tot_unemp + u 1743.136 0.106 0.
0563 0.0102 (1267.291 ) ( 0.
0273 ) (0.0014 ) ( 0.0065 ) t-statistic= 1.38 3.90 40.
88 1.56 t-critical= 2.33 2.
33 2.33 2.33 N= 3107 R-squared= 0.3787 Median household income and country population is statistically significant at 1% of significant level. But since t -statistic for total unemployment is less than critical value, it fails to reject the null hypothesis for this model.
total_allage = ?0 + ?1med_house + ?2population + ?3tot_unemp + ?4educ_less_high + u 1735.92 0.105 0.0562 0.0101 0.0053 (1 269.
222 ) ( 0.0274 ) ( 0.0014 ) ( 0.0065 ) ( 0.0024 ) t-statistic= 1.37 3.85 40.
68 1.55 2.23 t-critical= 2.33 2.
33 2.33 2.33 2.33 N= 3100 R-squared= 0.3797 Median household and country population is statistically significant at 1% of significant level.
But it fails to reject the null hypothesis for total unemployment and less than high school diploma at 1% of significant level. total_allage = ?0 + ?1med_house + ?2population + ?4educ_less_high + u 1750.61 0.106 0. 0562 0.005 4 (1268.
811 ) ( 0.0273 ) ( 0.0014 ) (0.0024 )t-statistic= 1.38 3.89 40.71 2.23 t-critical= 2.
33 2.33 2.33 2.33 N= 3102 R-squared= 0.3793 Median household income and country population are statistically significant at 1% of significant level .
But for the less than high school diploma, it fails to reject because t -statistic is less than critical value. For simple regression model, the R -squared is 0.0431 meaning s 4.31% of the variation in total number of people in poverty can be explained by the median household’s income. Independent variable did not explain much of variation in dependent variable. For four multiple regression model, they do not show much differences among each other. The highest R – squared is the third model which is 0.
3797. It means the regression explains 37.97% of the total variation in total number of people in poverty. The variation is a little bit low although it shows an increment compared to the simple model. It is because it included total unemployment variable in each model. In this journal stated that this explanator y variable has small correlation thus makes the model become weaken.
It might be the variable is insignificant variable or the model missed other significant variables that effect on poverty level in this country.
